A method for evaluating optical image quality of space targets
By combining median filtering, binarization, and closing operations with sharpness, size, intensity, and overexposure effect compensation coefficients, the problem of inaccurate evaluation of optical image quality of space targets in existing technologies is solved, and more accurate image quality assessment is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XI AN JIAOTONG UNIV
- Filing Date
- 2026-03-13
- Publication Date
- 2026-06-05
Smart Images
Figure CN122156771A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of image quality assessment technology, and in particular to a method for assessing the quality of optical images of space targets. Background Technology
[0002] Optical images of space targets are an important basis for identifying target components. However, under extreme imaging conditions, the optical images of space targets acquired by observation suffer from severe degradation coupling effects, which can easily lead to the degradation of target structural information. This poses great difficulties for image interpretation and target analysis, and seriously affects the effectiveness of space target component identification. Therefore, it is necessary to evaluate and classify the quality of a large number of space target images and select the image data with better quality as the object of target analysis.
[0003] Existing image quality assessment methods are all based on statistical information from image grayscale data. However, these methods are almost all derived from studies of existing image quality assessment standard datasets, which typically use objects such as scenes, people, and buildings as targets, significantly different from images of space targets. Space target optical imaging differs from conventional ground environment imaging and Earth remote sensing identification. The imaging environment is extremely unique, representing a target information acquisition process under extreme imaging conditions, characterized by extremely dark backgrounds, locally visible targets, and strong shadow effects, and the observation process is highly dynamic. Existing image quality assessment methods do not specifically consider the characteristics of space targets and the various effects under extreme space imaging conditions, nor do they evaluate and calculate the factors affecting the acquisition of space target image information. Therefore, these algorithms are not well-suited for space target optical image data and cannot meet the image selection requirements for space target component identification applications. Summary of the Invention
[0004] Therefore, it is necessary to provide a method for evaluating the optical image quality of space targets to address the aforementioned technical problems.
[0005] The following technical solution is adopted in this specification: This specification provides a method for evaluating the quality of optical images of space targets, including: The optical image of the space target is subjected to median filtering, and the filtered optical image of the space target is binarized by segmenting gray level thresholds. The foreground region of the image obtained after binarization is determined as the initial target region. The initial target region is processed by a closing operation to close the finely connected regions within the initial target region, thus obtaining the target region. The system acquires a sharpness score for optical images of space targets and determines the size, intensity, and overexposure compensation coefficient of targets in the optical images based on the target region. The overexposure compensation coefficient is used to deduct different degrees from the image quality score based on the severity of overexposure in the target region. The quality score of the optical image of the space target is determined based on the sharpness score, the size and intensity of the target, and the overexposure effect compensation coefficient.
[0006] Optionally, the method for determining the grayscale threshold includes: Detect the number of pixels in overexposed areas in optical images of spatial targets; Based on the relationship between the optimal segmentation grayscale threshold and the number of pixels in the overexposed area of the image, and the number of pixels in the overexposed area of the optical image of the spatial target, the segmentation grayscale threshold of the optical image of the spatial target is determined.
[0007] Optionally, the relationship between the optimal grayscale threshold for segmentation and the number of pixels in the overexposed region of the image is as follows: in, To determine the optimal grayscale threshold, This represents the number of pixels in the overexposed area of the image.
[0008] Optionally, the size of the target in the optical image of the spatial target is determined based on the target region, including: Find the minimum bounding rectangle of the target region; The size of the target is determined by the number of pixels in the smallest bounding rectangle of the target region.
[0009] Optionally, based on the target region, the intensity of the target in the optical image of the spatial target is determined, including: The average gray value of all pixels in the target area is used to determine the intensity of the target.
[0010] Optionally, based on the target region, an overexposure compensation coefficient for the target in the optical image of the spatial target is determined, including: Obtain the number of pixels in the overexposed area of the target region; Calculate the ratio of the number of pixels in the overexposed area to the total number of pixels in the target area; Based on this ratio, the overexposure effect compensation coefficient of the target is determined.
[0011] Optionally, the overexposure effect compensation coefficient of the target. The calculation formula is: in, This represents the ratio of the number of pixels in the overexposed area to the total number of pixels in the target area.
[0012] Optionally, a sharpness score is obtained for the optical image of the space target, including: Obtain the modulation transfer function (MTF) curve corresponding to the optical image of the space target; Based on the MTF curve, the MTF50 value is determined, and the MTF50 value is used as the sharpness score of the optical image of the space target.
[0013] Optionally, quality scoring of optical images of space targets. The calculation formula is: in, The size of the target in the optical image of the space target. The intensity of the target in the optical image of the space target. Scoring the sharpness of optical images of space targets. This is the overexposure compensation coefficient for the target in the optical image of the space target. , , , These represent the degree of influence of the corresponding factors on the overall image quality.
[0014] Optionally, , , , This specification provides a space target optical image quality evaluation device, comprising:
[0015] The extraction module is used to perform median filtering on the optical image of the spatial target, and to perform binarization on the filtered optical image of the spatial target by segmenting grayscale thresholds. The foreground region of the image obtained after binarization is determined as the initial target region. The initial target region is then processed by a closing operation to close the finely connected regions in the initial target region together to obtain the target region. The calculation module is used to obtain the sharpness score of the optical image of the space target, and to determine the size, intensity and overexposure effect compensation coefficient of the target in the optical image of the space target based on the target area; the overexposure effect compensation coefficient is used to deduct the image quality score to different degrees according to the severity of the overexposure effect in the target area. The determination module is used to determine the quality score of the optical image of a space target based on the sharpness score, the size and intensity of the target, and the overexposure effect compensation coefficient.
[0016] This specification provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for evaluating the optical image quality of space targets.
[0017] This specification provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for evaluating the optical image quality of space targets.
[0018] The above-mentioned technical solutions adopted in this specification can achieve the following beneficial effects: The space target optical image quality evaluation method provided in this specification extracts the initial target region of the image through median filtering and binarization, effectively eliminating noise and artifacts. It then fills in small holes within the target region using closing operations and connects adjacent minor breaks, restoring the target region to a complete and continuous shape. Furthermore, it quantitatively evaluates the target in the image based on four factors: the sharpness score of the space target optical image, the target size, intensity, and overexposure compensation coefficient. The combined evaluation results of size, intensity, sharpness score, and overexposure effect are used as the final quality score of the space target optical image. This method improves the evaluation accuracy of space target optical images. Attached Figure Description
[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:
[0020] Figure 1 This specification provides a flowchart illustrating a method for evaluating the quality of optical images of space targets. Figure 2 This is a schematic diagram illustrating a problem with the optical image quality of a space target, as provided in this specification. Figure 3 This diagram illustrates the initial target region extraction effect under different segmentation grayscale thresholds provided in this specification. Figure 4 This is a schematic diagram of the halo artifact effect in an optical image of a space target, provided in this specification. Figure 5 This specification provides a schematic diagram of the fitted straight line between the number of overexposed pixels and the optimal segmentation grayscale threshold. Figure 6 Images of spatial targets of different sizes provided in this specification; Figure 7 This document provides schematic diagrams of target images at different scales and their minimum bounding rectangle regions. Figure 8 This specification provides a curve of overexposure effect compensation coefficient; Figure 9 This is a schematic diagram of the optical image simulation of the space target provided in this specification. Detailed Implementation
[0021] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments in this specification without creative effort are within the scope of protection of this application.
[0022] Existing image quality assessment methods can be divided into subjective and objective methods. Subjective assessment involves directly observing the image using human visual perception and then evaluating and scoring the image quality according to one's own standards. Objective assessment methods establish mathematical models based on the human visual system and calculate image quality using specific formulas. They are characterized by batch processing capability, reproducible results, and are not susceptible to bias due to human error. Objective assessment methods are typically categorized into reference image assessment methods and referenceless image assessment methods based on their dependence on the original image. Key reference image assessment metrics include mean squared error, signal-to-noise ratio (SNR), peak signal-to-noise ratio (PSNR), improved SNR, and structural similarity metrics. Key referenceless image assessment metrics include average gradient and information entropy.
[0023] Existing image quality assessment methods do not specifically consider the degradation coupling effect under extreme space imaging conditions, and are not very applicable to optical image data of space targets, making it difficult to meet the needs of image quality assessment and screening in space target component identification scenarios.
[0024] Based on this, and addressing the need for image quality evaluation and selection in space target component recognition applications, this invention proposes a method for evaluating the optical image quality of space targets. Starting with an analysis of the optical imaging characteristics and image degradation features of space targets, it identifies target size, intensity, sharpness, and overexposure as four factors affecting the optical image quality of space targets. Quantitative calculation methods and parameter weights for the influence of each factor on image quality are provided, and the mathematical expression of the space target optical image quality evaluation model is determined. The evaluation effect of this method is verified based on typical space target optical image data of different qualities. The results show that the calculation results of this method are basically consistent with the subjective visual evaluation results, accurately reflecting the quality of space target images. This provides an effective approach for optimizing image data in space target component recognition applications.
[0025] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.
[0026] Figure 1 This is a schematic diagram of a method for evaluating the optical image quality of a space target as described in this specification, which specifically includes the following steps: S101, perform median filtering on the optical image of the spatial target, and binarize the filtered optical image of the spatial target by segmenting grayscale thresholds. The foreground region of the image obtained after binarization is determined as the initial target region. The initial target region is processed by closing operation to close the finely connected regions in the initial target region together to obtain the target region.
[0027] First, the characteristics of optical imaging of space targets are analyzed: optical imaging data of space targets have the following characteristics: First, the background is extremely dark. Except for the target area, the image is a black deep space background with no complex background environment interference, and the target position detection and identification is not difficult. Second, the image resolution is affected by atmospheric effects and imaging distance, and the ability to identify the details of target components is limited. Third, with sunlight as the only light source, the target brightness distribution is severely affected by the unidirectional light direction. The target's front side is brighter and the back side is very dark. Under the side illumination of sunlight, the target may be partially visible, and the dynamic range is very large. Fourth, under strong light conditions, the overexposure effect is severe, which leads to image grayscale saturation, making it impossible to identify the target's structural features and causing halo artifacts.
[0028] The main factors affecting the identification of optical image components of space targets are four aspects: dimness, blurriness, small size, and overexposure effect. Figure 2 As shown, Figure 2 The diagrams illustrate the optical image quality issues of space targets. (a) shows a weak target, (b) shows a blurred image, (c) shows a small target size, and (d) shows an overexposure effect. The weakness is due to factors such as insufficient scattered light from the target and severe atmospheric extinction, resulting in low pixel grayscale values in the target area and weak contrast between the target area and the background, making it difficult to accurately identify the target from the image. The blurriness is caused by atmospheric scattering and turbulence effects on the reflected light from the target, resulting in insufficient clarity of the target and its component structures in the image, a small grayscale gradient of the target edge contour, and difficulty in confirming the edge position. The small size is due to factors such as the small size of the target body and the long imaging distance, resulting in a small number of pixels in the target area, making it difficult to effectively identify the target's structural details from the image. The overexposure effect is due to excessive intensity of reflected light from the target, exceeding the upper limit of the camera's imaging device's exposure response range, causing excessive brightness and grayscale saturation in some target areas, leading to the loss of target structural information in the bright areas.
[0029] Image target region extraction is fundamental to quality analysis and evaluation. Target region extraction in optical images of spatial targets requires attention to three aspects: First, dark areas such as solar panel grids are prone to forming "holes" and being identified as background areas; second, narrow areas between target components are easily extracted and identified as target areas; third, overexposure-induced bright areas of the target will produce bright halo artifacts in the surrounding background area, easily identified as target areas. To address these characteristics, image morphology processing methods are used to extract spatial target regions in three steps: First, median filtering is used to suppress noise in the image to eliminate potential pixel noise in the background; then, an appropriate segmentation grayscale threshold is selected for image binarization to extract the initial target region. During this process, halo artifacts caused by overexposure should be avoided from being identified as initial target regions; finally, the initial target region is processed through a closing operation (dilation followed by erosion) to close the finely connected areas together, obtaining the target region. This aims to avoid target region separation under poor imaging conditions, preventing the target from being enclosed into multiple regions, and to prevent small dark areas of the target from being identified as background areas.
[0030] In one embodiment, such as Figure 3 As shown, Figure 3 The figures illustrate the initial target region extraction results under different segmentation grayscale thresholds. (a) shows the original spatial target optical image; (b) shows the effect of a segmentation grayscale threshold of 10; (c) shows the effect of a segmentation grayscale threshold of 25; and (d) shows the effect of a segmentation grayscale threshold of 50. It is evident that the key and challenge in accurately extracting spatial target regions lies in determining a suitable segmentation grayscale threshold and avoiding misidentifying halo artifacts caused by overexposure as targets. Figure 4 As shown, Figure 4 This is a schematic diagram of the halo artifact effect in optical images of space targets. The intensity of the halo artifact effect is positively correlated with the overexposure effect; that is, the more severe the overexposure effect, the stronger the halo artifact effect and the larger the gray value in the image.
[0031] Considering the relationship between halo artifacts and overexposure, the optimal segmentation grayscale threshold for each image can be determined by adjusting the target's segmentation grayscale threshold and analyzing the relationship between the optimal segmentation grayscale threshold and the number of overexposed pixels in the image. The results of this invention's analysis of multiple spatial target optical images with different overexposure levels are as follows:
[0032] Table 1. Correspondence between the number of pixels in the extremely bright region and the optimal segmentation grayscale threshold like Figure 5 As shown, Figure 5This is a schematic diagram of the fitted straight line between the number of overexposed pixels and the optimal segmentation grayscale threshold, from... Figure 5 It can be seen that there is a linear positive correlation between the number of pixels in the extremely bright region caused by overexposure in the spatial target image and the optimal grayscale threshold for target segmentation. A straight line with a slope of 0.016 achieves the best fit (image size is 480×480). The relationship between the optimal grayscale threshold for segmentation and the number of pixels in the overexposure region of the image is as follows:
[0033] (1) in, To determine the optimal grayscale threshold, This represents the number of pixels in the overexposed area of the image.
[0034] Therefore, based on the linear relationship between the number of pixels in overexposed areas of an image and the optimal segmentation threshold, the number of pixels in extremely bright areas can be statistically analyzed. The optimal grayscale threshold for target region segmentation is determined by calculation. .
[0035] Specifically, the method for determining the segmentation grayscale threshold includes: detecting the number of pixels in the overexposed area of the spatial target optical image; and determining the segmentation grayscale threshold of the spatial target optical image based on the relationship between the optimal segmentation grayscale threshold and the number of pixels in the overexposed area of the image, as well as the number of pixels in the overexposed area of the spatial target optical image.
[0036] S102, acquire the sharpness score of the optical image of the space target, and determine the size, intensity and overexposure effect compensation coefficient of the target in the optical image of the space target based on the target area; the overexposure effect compensation coefficient is used to deduct the image quality score to different degrees according to the severity of the overexposure effect in the target area.
[0037] When analyzing and evaluating image quality, it is necessary to analyze the factors affecting image quality and their evaluation methods based on the optical imaging conditions of the space target and the characteristics of the image. This invention sets four factors affecting the image quality of space targets: target size, intensity, sharpness, and overexposure effect. The following is an analysis and quantitative calculation of these factors.
[0038] Target size: For the size of a target, given a fixed image size, it can be represented by the number of pixels in the target region (its proportion in the image). There are two evaluation methods: the target region pixel count method and the target minimum bounding rectangle pixel count method.
[0039] (1) Target region pixel count method The target region pixel count method uses the number of pixels within the target region to characterize the target size. First, the image is processed by median filtering to eliminate the influence of pixel noise that may be present in the background; then, image grayscale calculation and target segmentation are performed to calculate the number of pixels occupied by the target region.
[0040] like Figure 6 As shown, Figure 6 Images of spatial targets at different scales are provided, with an image size of 480×480 and a total of 230,400 pixels. In the three images showing smaller, medium, and larger targets, the number of pixels in the target region is 7,140, 33,301, and 44,559, respectively, accounting for 3.10%, 14.45%, and 19.34% of all pixels. The number of pixels in the target region or its proportion can provide a relatively intuitive and accurate reflection of the target's size in the image.
[0041] (2) Minimum Target Bounding Rectangle Pixel Count Method The minimum bounding rectangle pixel count method characterizes the target size using the number of pixels occupied by the minimum bounding rectangle region of the target. Therefore, in one embodiment, determining the size of a target in a spatial target optical image based on the target region includes: obtaining the minimum bounding rectangle of the target region; and determining the number of pixels in the minimum bounding rectangle of the target region as the target size. For example... Figure 7 As shown, Figure 7 This is a schematic diagram of target images at different scales and their minimum bounding rectangle regions.
[0042] Table 2. Calculation results of the target region pixel number method and the target minimum bounding area method for targets of different scales. The two methods described above are similar in their effectiveness in assessing target size. However, the minimum bounding rectangle pixel count method is theoretically superior because there may be some dark areas (which actually belong to the target area) in the spatial target region. The first method is likely to identify these areas as background areas, while the second method avoids this problem and obtains a target area that is closer to the actual situation.
[0043] Target strength: The intensity characteristics of a target can be characterized by the grayscale distribution of pixels in the target region, with the average grayscale value of the target region used as a quantitative evaluation index. Therefore, in one embodiment, determining the intensity of a target in a spatial target optical image based on the target region includes: determining the average grayscale value of all pixels in the target region as the intensity of the target.
[0044] Average gray value of all pixels in the target area The expression is as follows: (2) in, This represents the number of pixels in the target area. Represents any cell in the target region. Represents a cell The grayscale value.
[0045] As shown in Table 3, the grayscale distribution and average grayscale value of the target region at different intensities are as follows. It can be seen that the stronger the target brightness in the image, the more concentrated the grayscale distribution of the target region is in the higher grayscale area, and the larger the average grayscale value of the target; conversely, the weaker the target brightness in the image, the more the grayscale distribution of the target region tends to be in the lower grayscale area, and the smaller the average grayscale value of the target. This also conforms to the visual characteristics of human eye observation of images.
[0046] Table 3. Gray-scale distribution and average gray-scale value of target areas with different brightness levels. Clarity of objectives: Target sharpness can be evaluated using a sharpness rating system for optical images of space targets. The richness of detail information, such as the edge contours of the structure and components of a space target, is a crucial factor in evaluating the image quality. High image sharpness indicates rich target feature information, while a blurry image suggests a significant loss of target information. Currently, there are three main methods for evaluating image sharpness: no-reference evaluation, full-reference evaluation, and half-reference evaluation. Since a clear reference image of a space target is generally unavailable, the no-reference model method is chosen for evaluating the sharpness of space target images.
[0047] The Modulation Transfer Function (MTF) is a widely used image sharpness evaluation method in computer vision and image processing. It calculates the blurriness and sharpness of edges in an image to obtain the MTF curve. MTF50 is the frequency (Cycle Per Pixel) at which the MTF curve value drops to 50% of its maximum value, and it is a widely used standard for measuring sharpness. This invention uses MTF50 as an image sharpness evaluation index. Table 4 shows the MTF curves and MTF50 values of two spatial targets with different sharpness.
[0048] Table 4. MTF curves and MTF50 values of spatial target images at different resolutions. As can be seen from the table above, the MTF50 value of an image can accurately describe and characterize the sharpness (blurriness) of a spatial target image, and can be used as a quantitative evaluation index of the sharpness of a spatial target image.
[0049] Therefore, obtaining the sharpness score of the optical image of the space target includes: obtaining the modulation transfer function (MTF) curve corresponding to the optical image of the space target; determining the MTF50 value based on the MTF curve, and using the MTF50 value as the sharpness score of the optical image of the space target.
[0050] Overexposure effect of the target: If overexposure occurs due to excessively strong reflected light from the target, resulting in a large number of gray-saturated areas within the target, the target intensity score will be high. However, the target's shape and structural information will be "submerged" in these gray-saturated pixels, severely impacting the extraction of target structural features. Therefore, overexposure needs to be penalized in image quality evaluation. If the proportion of oversaturated gray areas in the target area is too high, the quality evaluation score will be reduced to a certain extent; the higher the proportion of oversaturated areas, the greater the penalty. It's easy to see that if all pixels in the target area are oversaturated, the image contains very little information, although it may still contain basic outline information of the target. Therefore, a small amount of evaluation score should be retained. Based on this idea, this invention proposes an overexposure compensation coefficient. As a factor in the image quality evaluation score expression, as shown in formula (3), where R is the ratio of the number of pixels in the overexposed area to the number of pixels in the target area. An overexposure effect compensation coefficient is set. The expression is in piecewise form, as follows:
[0051] (3) In one embodiment, determining the overexposure effect compensation coefficient of a target in a spatial target optical image based on the target region includes: obtaining the number of pixels in the overexposure region of the target region; calculating the ratio of the number of pixels in the overexposure region of the target region to the number of pixels in the target region; and determining the overexposure effect compensation coefficient of the target based on the formula (3) according to the ratio.
[0052] like Figure 8 As shown, Figure 8 The graph shows the overexposure effect compensation coefficient curve. As can be seen from the graph, the overexposure effect compensation coefficient... The value is between 0 and 1, indicating the degree of overexposure effect on image quality: when the overexposed area accounts for less than 20% of the target area, the overexposure effect is weak. A value of 1 has no impact on image quality; however, as the proportion of overexposed areas increases, the overexposure effect intensifies. As the value decreases, the impact on image quality gradually increases; if the entire target area is overexposed, then... The value is only 0.1.
[0053] The table below shows the pixel proportion of overexposed areas and the overexposure compensation coefficient for typical spatial target images with four different levels of overexposure: extremely strong, strong, moderate, and weak. Calculation results. It can be seen that the overexposure effect compensation coefficient... It can deduct different degrees from the image quality score based on the severity of the overexposure effect; for images with a small overexposure effect, the overexposure effect compensation coefficient is adjusted. A value of 1 indicates that the overexposure effect has no impact on image quality and does not change the image quality score.
[0054] Table 5 Typical spatial target images and overexposure compensation coefficients S103 determines the quality score of the optical image of the space target based on the sharpness score, the size and intensity of the target, and the overexposure effect compensation coefficient.
[0055] Based on the above four factors affecting image quality—size, intensity, sharpness score, and overexposure effect—and their calculation methods, this invention sets weights by combining subjective and objective evaluation methods, and designs and establishes an image quality evaluation model suitable for optical images of space targets by integrating various influencing factors.
[0056] The quality score of an optical image of a space target is determined by the product of the scores of four factors: the size, intensity, sharpness, and overexposure compensation coefficient of the target in the image, as shown in formula (4). The calculation formula is:
[0057] (4) in, The size of the target in the optical image of the space target. The intensity of the target in the optical image of the space target. Scoring the sharpness of optical images of space targets. This represents the overexposure compensation coefficient for targets in optical images of space targets. Since the effects of these factors on image quality are independent, they are multiplied together. , , , These represent the degree of influence of the corresponding factors on the overall image quality.
[0058] When determining the indices for the scores of each factor, consideration was given to different image quality conditions. The values range from hundreds to tens of thousands. The values generally range from 1 to 200, with a maximum difference of over a hundred times; while Values under image sharpness and blurriness conditions, and The values can differ by a factor of approximately 10 under different overexposure levels. To ensure a balanced impact of various factors on overall image quality, a set parameter is used in the comprehensive evaluation of spatial target image quality. and index , ,set up and index , The quality score of the optical image of the space target. The expression is as follows:
[0059] (5) In one embodiment, to verify the accuracy and effectiveness of the image quality assessment model proposed in this invention, multiple simulated optical images of spatial targets of different qualities under conditions of different target sizes, intensities, sharpness, and overexposure effects were selected. The method provided by this invention was used for calculation to verify the evaluation effect of the method. Some of the images and their quality assessment results are shown below. Figure 9 As shown in Table 6.
[0060] Table 6. Calculation Results of Image Quality Analysis and Evaluation for Typical Spatial Targets Based on subjective perceptions of image quality and comparing the above image quality evaluation calculation results, it can be seen that the invention proposed in this paper and its calculation results are basically consistent with the subjective evaluation results, and can accurately reflect the image quality of space targets, and can be adapted to the evaluation of optical image quality of space targets.
[0061] Based on comprehensive image quality evaluation indicators The calculation results show that images with higher quality levels... The value is larger, therefore it can be used to evaluate the overall image quality index. The image quality is automatically graded according to a certain threshold. For the purpose of spatial target component recognition, image quality is divided into three levels: excellent, medium, and poor. Excellent quality images contain clearly identifiable target components with rich feature information and can be used as datasets for spatial target component detection and recognition. Medium quality images have generally visible overall target outlines but limited structural feature information and are generally not included in the recognition target set; they can be considered for use when the number of excellent quality images is insufficient. Poor quality images contain targets that are too small or whose structures are difficult to identify and are not considered as data sources for processing and analysis.
[0062] The space target optical image quality evaluation method proposed in this invention automatically grades the quality of image data, requiring the determination of evaluation score thresholds for each quality level. According to... Figure 9 The nine typical optical images of space targets shown and their quality evaluation results can be analyzed according to the comprehensive image quality evaluation index. The value categorizes the image into three quality levels: Images with a value greater than 1.0E+10 are considered to have excellent quality. Image quality is medium for values between 1.0E+08 and 1.0E+10. Images with a value less than 1.0E+08 are considered poor quality. That is, images 6 and 8 are of excellent quality, images 1, 2, 3, and 5 are of medium quality, and images 4, 7, and 9 are of poor quality.
[0063] It is evident that the space target optical image quality evaluation method and quality grading threshold proposed in this invention enable automatic quality evaluation and grading of space target optical images, select image data that can be used for space target analysis and recognition from a large amount of image data, construct a high-quality space target component recognition image database, and provide important data support for the training and testing of component recognition algorithms.
[0064] The execution subject of the method provided by this invention can be a server, which can be a server set up on a business platform, or a device such as a desktop computer or laptop computer that can execute the solution in this specification.
[0065] When applying the space target optical image quality evaluation method provided in this manual, it is not necessary to consider... Figure 1 The steps shown are executed in sequence. The specific execution order of each step can be determined as needed, and this manual does not impose any restrictions on it.
[0066] The above describes one or more embodiments of a space target optical image quality evaluation method provided in this specification. Based on the same idea, this specification also provides a corresponding space target optical image quality evaluation device, which includes: The extraction module is used to perform median filtering on the optical image of the spatial target, and to perform binarization on the filtered optical image of the spatial target by segmenting grayscale thresholds. The foreground region of the image obtained after binarization is determined as the initial target region. The initial target region is then processed by a closing operation to close the finely connected regions in the initial target region together to obtain the target region. The calculation module is used to obtain the sharpness score of the optical image of the space target, and to determine the size, intensity and overexposure effect compensation coefficient of the target in the optical image of the space target based on the target area; the overexposure effect compensation coefficient is used to deduct the image quality score to different degrees according to the severity of the overexposure effect in the target area. The determination module is used to determine the quality score of the optical image of a space target based on the sharpness score, the size and intensity of the target, and the overexposure effect compensation coefficient.
[0067] Specific limitations regarding the space target optical image quality assessment device can be found in the limitations of the space target optical image quality assessment method described above, and will not be repeated here. Each module in the aforementioned space target optical image quality assessment device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.
[0068] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 A method for evaluating the quality of optical images of space targets is provided.
[0069] This specification also provides a computer device, which, at the hardware level, includes a processor, an internal bus, a network interface, memory, and non-volatile memory, and may also include other hardware required for business operations. The processor reads the corresponding computer program from the non-volatile memory into memory and then executes it to achieve the above-mentioned functions. Figure 1 A method for evaluating the quality of optical images of space targets is provided.
[0070] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.
[0071] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
Claims
1. A method for evaluating the quality of optical images of space targets, characterized in that, include: The optical image of the space target is subjected to median filtering, and the filtered optical image of the space target is binarized by segmenting gray level thresholds. The foreground region of the image obtained after binarization is determined as the initial target region. The initial target region is processed by a closing operation to close the finely connected regions within the initial target region, thus obtaining the target region. The system acquires a sharpness score for optical images of space targets and determines the size, intensity, and overexposure compensation coefficient of targets in the optical images based on the target region. The overexposure compensation coefficient is used to deduct different degrees from the image quality score based on the severity of overexposure in the target region. The quality score of the optical image of the space target is determined based on the sharpness score, the size and intensity of the target, and the overexposure effect compensation coefficient.
2. The method according to claim 1, characterized in that, The methods for determining the grayscale threshold include: Detect the number of pixels in overexposed areas in optical images of spatial targets; Based on the relationship between the optimal segmentation grayscale threshold and the number of pixels in the overexposed area of the image, and the number of pixels in the overexposed area of the optical image of the spatial target, the segmentation grayscale threshold of the optical image of the spatial target is determined.
3. The method according to claim 2, characterized in that, The relationship between the optimal grayscale threshold for segmentation and the number of pixels in the overexposed region of the image is as follows: in, To determine the optimal grayscale threshold, This represents the number of pixels in the overexposed area of the image.
4. The method according to claim 1, characterized in that, Based on the target region, determine the size of the target in the optical image of the space target, including: Find the minimum bounding rectangle of the target region; The size of the target is determined by the number of pixels in the smallest bounding rectangle of the target region.
5. The method according to claim 1, characterized in that, Based on the target region, determine the intensity of the target in the optical image of the space target, including: The average gray value of all pixels in the target area is used to determine the intensity of the target.
6. The method according to claim 1, characterized in that, Based on the target region, determine the overexposure compensation coefficient for the target in the optical image of the spatial target, including: Obtain the number of pixels in the overexposed area of the target region; Calculate the ratio of the number of pixels in the overexposed area to the total number of pixels in the target area; Based on this ratio, the overexposure effect compensation coefficient of the target is determined.
7. The method according to claim 6, characterized in that, Overexposure effect compensation coefficient of the target The calculation formula is: in, This represents the ratio of the number of pixels in the overexposed area to the total number of pixels in the target area.
8. The method according to claim 1, characterized in that, Obtain a sharpness score for optical images of space targets, including: Obtain the modulation transfer function (MTF) curve corresponding to the optical image of the space target; Based on the MTF curve, the MTF50 value is determined, and the MTF50 value is used as the sharpness score of the optical image of the space target.
9. The method according to claim 1, characterized in that, Quality rating of optical images of space targets The calculation formula is: in, The size of the target in the optical image of the space target. The intensity of the target in the optical image of the space target. Scoring the sharpness of optical images of space targets. This is the overexposure compensation coefficient for the target in the optical image of the space target. , , , These represent the degree of influence of the corresponding factors on the overall image quality.
10. The method according to claim 9, characterized in that, ， ， ， 。